Dedifferentiated liposarcoma, from the amplicon up

What dedifferentiated liposarcoma is

DDLPS is not a tumour that starts from scratch. It develops out of an atypical lipomatous tumour / well-differentiated liposarcoma when part of that fatty tumour stops looking like fat and starts growing as a high-grade, non-lipogenic sarcoma.¹ Both components usually sit in the same mass, which is one reason a small biopsy can miss the aggressive part.

It grows most often in the retroperitoneum — the space behind the abdominal cavity, where a tumour can become very large before it causes symptoms. The dedifferentiated component is morphologically unpredictable: it can imitate undifferentiated pleomorphic sarcoma, solitary fibrous tumour, myxoid liposarcoma, and in some variants even carcinoma, melanoma or lymphoma. Preoperative biopsies in one published series of rhabdoid/epithelioid DDLPS never suggested the correct diagnosis.²

This is precisely why the molecular result matters more than the microscope in this family of tumours.

The amplicon is the diagnosis

Nearly all DDLPS carry high-level amplification of MDM2 and CDK4.¹ When a pathologist has to decide whether a retroperitoneal high-grade sarcoma is a DDLPS or something else, MDM2 FISH is generally the test that settles it. Other genes on the same 12q13-15 stretch — DDIT3, STAT6, GLI1, HMGA2 — are amplified along with it and explain some of the odd morphologies and misleading immunostains.³

The full explanation of what amplification means, how it is tested, and what it changes is on our page MDM2 and CDK4 amplification, explained.

Why this page exists on an intimal sarcoma site

Intimal sarcoma is defined by the same amplification.⁴¹⁰ The two diseases differ in where they grow, how they are operated on and how they kill — but they share the molecular engine that current targeted drugs are aimed at.

The practical consequence is asymmetric. DDLPS has enough patients to run a randomised phase III trial; intimal sarcoma, with a few hundred cases a year worldwide, does not and probably never will. So the evidence that exists for the driver was generated in DDLPS, and anyone with an MDM2/CDK4-amplified tumour has a reason to read it.

The limits are real and worth stating plainly. Evidence generated in one histology does not automatically transfer to another: the tumour microenvironment differs, the surgical situation differs, and the additional genomic alterations differ. Read-across is an argument for discussing a drug and looking for a trial — not a substitute for evidence in your own disease.

Treatment today

Surgery

Complete resection at a sarcoma reference centre is the only approach with curative intent, and the quality of that first operation is the single largest lever on long-term outcome. Retroperitoneal DDLPS recurs locally in a large share of patients, and re-operations become progressively harder.

Systemic therapy

The conventional sequence has been built on cytotoxic drugs: doxorubicin-based chemotherapy first, with ifosfamide, gemcitabine/docetaxel, eribulin used later. Eribulin has randomised data in advanced liposarcoma specifically.¹ A note on pazopanib: its registration trial in soft-tissue sarcoma excluded adipocytic tumours, so it is not a standard choice in liposarcoma — a distinction that is easy to lose when reading general sarcoma summaries.

CDK4/6 inhibition

A phase II trial of abemaciclib in 30 patients with progressing DDLPS reported median progression-free survival of 33 weeks, with 77% progression-free at 12 weeks.⁵ SARC041, a randomised double-blind phase III in 108 patients with advanced DDLPS, then compared abemaciclib against placebo:

Two things are worth taking from that table. The effect on progression is large and the effect on tumour size is small — this is a drug that holds disease still. And the overall-survival comparison is blurred because most placebo patients later received the drug anyway.⁶

MDM2 inhibition

Blocking MDM2 to release p53 has shown activity in early-phase trials,⁷ but the randomised MANTRA study of milademetan versus the standard comparator chemotherapy in DDLPS did not meet its primary endpoint.⁸ Further MDM2 inhibitors are in trials, some combined with immunotherapy.

Immunotherapy

Checkpoint inhibitors are being tested in DDLPS in combination with chemotherapy. Single-agent PD-1/PD-L1 blockade deserves a specific conversation, because MDM2 family amplification has been associated with unusually rapid progression on checkpoint monotherapy in a retrospective series.⁹

Trials

Because the driver is shared across diagnoses, some of the most interesting studies for a DDLPS patient are not DDLPS trials at all — they recruit on the amplification itself. Our current list, with enrolment status and links to the official registry entries, is at trials that enrol on MDM2 or CDK4 amplification.

Frequently asked